Page 1/22

Transcriptional pro�ling of gametogenesis in thegreen seaweed Ulva mutabilis identi�es an RWP‐RKtranscription factors linked to reproductionXiaojie Liu 

Ghent UniversityJonas Blomme 

Ghent UniversityKenny Bogaert 

Ghent UniversitySo�e D’hondt 

Ghent UniversityThomas Wichard 

Friedrich Schiller University JenaOlivier De Clerck  ( Olivier.declerck@ugent.be )

Ghent University

Research Article

Keywords: Ulva, green seaweeds, reproduction, gametogenesis, transcriptome, RWP-RK transcriptionfactor

Posted Date: August 25th, 2021

DOI: https://doi.org/10.21203/rs.3.rs-828026/v1

License: This work is licensed under a Creative Commons Attribution 4.0 International License.  Read Full License

Page 2/22

Abstract

BackgroundThe molecular mechanism underlying sexual reproduction in land plants is well understood in modelplants and is a target for crop improvement. However, unlike land plants, the genetic basis involved intriggering reproduction and gamete formation remains elusive in most seaweeds, which are increasinglyviewed as an alternative source of functional food and feedstock for energy applications.

ResultsGametogenesis of Ulva mutabilis, a model organism for green seaweeds, is studied. We analyzetranscriptome dynamics at different time points during gametogenesis following induction ofreproduction by fragmentation and removal of sporulation inhibitors. Analyses demonstrate that 45% ofthe genes in the genome are differentially expressed during gametogenesis. We identi�ed severaltranscription factors that potentially play a key role in the early gametogenesis of Ulva given the functionof their homologs in higher plants and microalgae. In particular, the detailed expression pattern of anevolutionary conserved transcription factor containing an RWP-RK domain suggests a key role duringUlva gametogenesis.

ConclusionsTranscriptomic analyses of gametogenesis in the green seaweed Ulva highlight the importance of aconserved RWP-RK transcription factor in induction of sexual reproduction. The identi�cation of putativemaster regulators of gametogenesis provides a starting point for further functional characterization.

BackgroundSexual reproduction is one of the most important and conserved processes in eukaryotes. In essence,sexual reproduction encompasses the fusion of two haploid gametes of opposite sex to form a diploidzygote. Zygote formation is either followed by meiosis to restore the haploid state (zygotic meiosis) or bydevelopment of a diploid life stage through mitotic divisions (gametic meiosis) [1]. Molecular geneticstudies of reproduction in land plants resulted in an excellent understanding of the fundamentalbiological principles and genes involved. However, unlike land plants, the molecular mechanisms involvedin the onset of reproduction, the formation of gametes (gametogenesis), fusion and meiosis remainelusive for most algae and seaweeds. In the green algal lineage, the molecular mechanisms underlyinggametogenesis and fusion have been extensively studied in Chlamydomonas reinhardtii, a unicellularfreshwater green alga [2–5], but to which extent key regulatory genes are conserved across green algae oreven land plants remains unclear.

Page 3/22

Here, we study the expression of genes during the various stages of gamete formation in Ulva, anemerging model for green seaweeds [6–8]. Ulva species are abundant in coastal benthic communitiesaround the world and form a potential source of biomass that can be used for food, feed, biofuel andnutraceuticals [8–13]. Given its commercial value, reproducibly controlling the life cycle of Ulva isimportant in the development of this crop. Under natural conditions, large parts of the Ulva thallus areconverted into gametangia or sporangia in mature individuals and the entire cell content is converted intoreproductive cells. Numerous studies have investigated the effect of different endogenous (e.g.production of sporulation and swarming inhibitors) and environmental factors (e.g. light, temperature ornutrient conditions) on gametogenesis in different Ulva species [14–18]. Although we are still farremoved from a complete understanding of the interplay of the various factors that appear to play a role,thallus fragmentation has been recognized as the most e�cient way to induce gametogenesis [15, 19–22].

The ability to e�ciently induce gametogenesis in laboratory cultures of Ulva mutabilis provides anelegant time course to study both phenotypic and molecular factors involved in this process [16]. Twosporulation inhibitors, SI-1 and SI-2 control the onset of gametogenesis of Ulva mutabilis. SI-1 is aspecies-speci�c glycoprotein secreted in the culture medium which depletes with maturation of thethallus. SI-2 is a nonprotein molecule existing within the inner space between the two blade cell layers.Thallus fragmentation followed by a washing step whereby spent culture medium is replaced, results inthe total removal of sporulation inhibitors and induces gametogenesis [15]. Following fragmentation andwashing, SI-1 and SI-2 are removed and vegetative cells enter a determination phase (~ 0-26h, Fig. 1) inwhich a 'swarming inhibitor' (SWI) is produced [16]. SWI functions as a mechanism to synchronizegamete release. The start of the differentiation phase coincides with the completion of the S-phase andthe G2-phase (~ 36 h) in the �rst cell cycle (Fig. 1). At the end of the determination phase, the cells enterthe next S-phase and become irreversibly committed to gametangium differentiation. The differentiationphase (~ 26-70h) is characterized by a reorientation of the chloroplast, followed by four consecutive celldivisions forming sixteen progametes per cell and their maturation. Mature gametes are eventuallyreleased during the swarming phase, following a light stimulus or if the SWI declines in concentration onthe third day after induction [16, 23]. Given that the whole process of Ulva gametogenesis, can besubdivided in discrete phases, the gene regulatory networks involved in the different stages ofgametogenesis should be distinct. As a �rst step toward understanding the molecular mechanisms ofgametogenesis, we established the transcriptional structure for Ulva mutabilis gametogenesis in thisresearch. We make use of differential removal of SI’s to separate the effect of fragmentation ontranscription from gametogenesis. Transcription patterns are compared with those of other green algaeand land plants.

ResultsTo dissect the molecular mechanisms underlying Ulva gametogenesis, the global gene expression levelsof U. mutabilis were measured by RNA-Seq as a function of time. We sampled �ve-time points: 0 h, 6 h,24 h, 48 h and 72 h after induction of gametogenesis (Fig. 1) under standard induction conditions (CW)

Page 4/22

and two-time points (6h and 24h after induction of gametogenesis) that omitted the washing step (CNW),with three biological replicates for each time point and condition. RNA-sequencing resulted in 215 x 106

reads, >70% mapped to the U. mutabilis genome. The total number of mapped reads for each sample andthe raw reads counts for each gene were listed in Table S2 and Table S3, respectively. Transcript pro�leswere highly reproducible among the three biological replicates at each time point based on the PearsonR2 test and PCA analysis (Table S4, Fig. S1). 

Gene expression pattern during gametogenesis

In total, 8296 distinct genes (62.2% of annotated genes) were expressed during gametogenesis, withrelatively low variation in the total number of expressed genes between the time points: ranging between7146 (0 h) and 7949 expressed genes (72 h). We identi�ed 6056 differentially expressed genes (DEGs)between any two of the �ve time points (0 h, 6 h, 24 h, 48 h, 72 h) during the gametogenesis process,representing 45% of genes of the annotated genome. Each time point could be characterized by a set ofgenes describing the various steps of gametogenesis (Fig. 1). Hierarchical clustering grouped genes in 5clusters for further analyses based on the stabilization of the SSE and the average silhouette widthvalues (Fig. S2). 

The identi�ed clusters were characterized by nearly unique sets of enriched GO terms and re�ected theidenti�ed phases of the gametogenesis process well (Fig. 2). Cluster 1 grouped genes expressed in thevegetative phase with decreasing expression levels during gametogenesis. Cluster 2 was the only clusterwith high expression during the determination phase but low expression during the differentiation andswarming phase. Cluster 3 and 4 grouped genes with high expression during the differentiation phase.The main difference between both clusters consisted of genes of cluster 3 being downregulated duringthe swarming phase. In contrast, genes in clusters 4 were characterized by a high expression level duringthe swarming phase also. Cluster 5 contains the least number of genes, upregulated only in the swarmingphase. 

DEGs related to photosynthesis and carboxylic acid biosynthesis were upregulated during thedetermination phase. Next, genes involved in DNA replication and microtubule-based movements have ahigh expression during the early differentiation process of gametogenesis as well as during the gameteformation. However, the different enriched GO terms between the two clusters apply primarily to DNAreplication which indicates that DNA replication is completed in the early phase of the differentiationprocess, before gametes formation starts. And the �nal swarming phase was mainly relevant to thepigment synthesis and cellular amino acid biosynthesis which was the �nal step for the gametesformation. 

Initiation of gametogenesis versus response to fragmentation

By adding an additional control group whereby thalli were fragmented but the sporulation inhibitors werenot removed (CNW, chopping no washing), we aimed to disentangle the effect of fragmentation on geneexpression from the induction of gametogenesis. By analyzing DEGs between CW and CNW group at 6 h

Page 5/22

and 24 h, we identi�ed 901 and 1137 signi�cantly differential expressed genes (FDR<0.05, log2FC>1 or<-1) at 6 h and 24 h, respectively. DEGs included 493 up-regulated and 408 down-regulated genes at 6 hand 652 up-regulated and 485 down-regulated genes at 24 h. Nearly half of the DEGs at 6 h have thesame expression pro�le at 24 h compared to the CNW group (Fig. 3). Of the 493 up-regulated and 408down-regulated genes at 6 h, 206 and 219 genes showed the same expression pro�les at 24 h, indicatingthat DEGs with a function in the determination phase tended to have continuous expression pro�lesduring early gametogenesis. The enriched GO terms for the up-regulated DEGs at 6 h are mainly related tothe photosynthesis, ATP biosynthetic process and oxidation-reduction processes (Fig. 4), which are inaccordance with cluster 2 enrichment. In contrast, GO terms signi�cantly enriched at 24 h are mainlyrelated to the microtubule-based movement, DNA replication and cilium organization, suggesting that theDNA replication initiation for the gametogenesis starts from 24 h, consistent with the expression patternof cluster 3 (Fig. 2). The over-represented GO terms for the down-regulated genes were related to theresponse to oxidative stress and oxidation-reduction process both at 6 h and 24 h after induction ofgametogenesis.

To identify the possible key initiator for Ulva gametogenesis, we analyzed the differentially expressed TFsat 6 h which signi�es the early responder to the washing treatment exclusively by comparing thechopping and washing (CW) group with the chopping without washing (CNW) group. We identi�ed 12 up-regulated and 6 down-regulated transcription factors (TFs) at 6 h (Table 1). Several homologs of theseTFs, as represented by the top blast hits in the Ulva-PLAZA database, are involved in gamete formation inArabidopsis or Chlamydomonas. 

Table 1. Differential expressed TFs of induced (CW) versus non-induced (CNW) thalli at 6 h. Log2FC>1:upregulated, LogFC<-1: downregulated. The homologs of the TFs of A. thaliana and C. reinhardtii arelisted in the "AtGID" and "CrID" columns respectively. Gene function annotation are summarized from TAIRor Phytozome databases.

Page 6/22

GeneID AtGID CrID Family/Domain Log2FC Gene functionannotation

UMSL012_0085 AT4G18770 CR03G00530 MYB 2.55 MYB98: expressedin the synergidcells, affect thefemalegametophytea

UMSL129_0009 N/A N/A MYB 1.85 N/A

UMSL039_0026 AT2G43770 CR12G01870 DEAD 2.21 FAP 52: �agellarassociated proteinb

UMSL054_0061 N/A N/A DEAD 2.81 N/A

UMSL003_0532 AT5G58080 CR02G02460 GARP_G2-like 1.08 ARR18: member ofresponse regulatora

UMSL103_0003 AT5G43990 N/A SET 1.53 SUVR2: regulationof eukaryotic geneexpression andchromatinstructurea 

UMSL008_0028 AT5G57390 CR01G02250 AP2/EREBP 1.49 AIL5: involved ingermination andseedling growtha

UMSL032_0028 AT2G01830 CR06G00550 DEAD 1.08 AHK4: cytokinin-binding receptorthat transducescytokinin signalsa

UMSL057_0048 AT5G53040 CR03G09530 RWP-RK 1.62 GR: promoteszygote elongationand basal cellfatesa

UMSL024_0043 AT3G28730 CR12G13130 HMG 4.77 ATHMG: binds tothe promoter ofrepressor of�owering: FLCa

UMSL033_0026 AT2G25170 CR03G03890 SWI/SNF_SNF2 1.51 CHD3: involved inpost-germinationrepression ofembryonicdevelopmenta

UMSL003_0659 AT3G27730 CR10G06300 DEAD 1.52 MER3: DNAhelicase requiredfor interference-sensitive meioticcrossovera

Page 7/22

UMSL151_0016 AT3G16770 CR16G05640 AP2 -1.32 ATEBP: ethyleneresponse factora

UMSL085_0029 AT2G17520 CR08G03590 zf-CCCH -1.27 ATIRE1-2: involvedin the regulation ofthe ER stressresponsive genesa

UMSL051_0112 AT5G41370 CR06G12570 DEAD -1.14 ATXPB1: encodes aDNA repair proteina

UMSL003_0378 AT1G19270 CR10G01560 DEAD -1.11 DA1: controls theinitiation of axillarymeristemsa

UMSL053_0033 AT3G58680 CR02G09000 MBF1 -1.03 FAP280: �agellarassociated proteinb

UMSL011_0269 AT1G01040 N/A DEAD -1.03 ABNORMALSUSPENSOR 1:encodes a Dicerhomologa

a Access from TAIR database (https://www.arabidopsis.org/)

b Access from Phytozome database (https://phytozome.jgi.doe.gov/pz/portal.html)

We acquired the homologs of the Arabidopsis and Chlamydomonas of the upregulated TFs at 6 h byblast. With reported function of the homologs, we tried to �nd the conserved TFs function in plantsgametogenesis. Particularly, an extensive studied TF containing RWP-RK domain aroused our interestsfor further characterization. The RWP-RK gene in Chlamydomonas, which is minus dominance (MID) isresponsible for switching on the minus-programme and switching off the plus-programme in gametedifferentiation [2]. Studies in Arabidopsis indicate that the RWP-RK TFs control cell differentiation duringfemale gametophyte development [24]. We investigated the expression pattern of the RWP-RK TF(UMSL057_0048) in more detail by means of qRT-PCR analysis, sampling tissue fragments undergoinggametogenesis every 2 hours. At each timepoints the expression was compared against a controltreatment not undergoing gametogenesis. UMSL057_0048 is up-regulated in the early determinationphase and throughout the determination and differentiation phase with maximum expression levels at 6h (Fig. 5a). More speci�cally, UMSL057_0048 shows a quick response to the chopping and washingtreatment with slight upregulation at 2 h after induction and signi�cantly upregulation at 4 h. 

In an additional experiment we contrasted the expression at 6 h of UMSL057_0048 to fragmented thallifrom which the SI were not removed (CNW) and unfragmented thalli that were washed to remove the SI inthe medium and hence were partially induced (NCW). qRT-PCR results indicate UMSL057_0048 is notupregulated at 6 h in treatments where gametogenesis is not induced (CNW and NCNW). In contrast inboth treatments where gametogenesis is induced (CW and NCW), UMSL057_0048 is signi�cantly

Page 8/22

upregulated (Fig. 5b). The Ulva genome contains one additional gene with an RWP-RK domain,UMSL048_0014. The latter, however, is not expressed (CPM value ~0) during gametogenesis. 

DiscussionThe transition from a vegetative to a reproductive state is a key step in the life cycle of Ulva. Analysis ofthe molecular mechanisms underpinning this process is pivotal for e�cient seeding technologies andputative genetic improvement for Ulva as a commercial crop species. In this study, RNA-seq was appliedto analyze the dynamic changes of gene expression of Ulva mutabilis and to identify key TFs involved inthe initiation of Ulva gametogenesis. Transcriptional analysis showed that nearly half (45.4%) of theprotein-coding genes predicted in the draft genome are differential expressed between at least 2timepoints during gametogenesis. The large reprogramming of the transcriptome likely re�ects the fulltransformation of a vegetative cell into 16 gametes [16].

Dynamic changes in gene regulation during gametogenesisGO enrichment analysis of DEGs clusters sheds light on the most enriched up- or down-regulatedbiological process during each of the three phases that characterize gametogenesis in Ulva. Cluster 1contains DEGs that are downregulated during the determination phase and genes in this cluster areenriched in the cell adhesion and pentose-phosphate shunt activity. For cluster 2 which is upregulatedduring the determination phase, there is a high expression of genes that are associated withphotosynthesis, carboxylic acid biosynthesis and ribose phosphate metabolism. This correlates withsigni�cant increases in maltose during gametogenesis of U. prolifera at 12 h and 24 h is, which areapproximately 12-fold change up-regulated compared to the control group [25]. Similarly, the upregulatedphotosynthesis related genes at 24 h and 48 h may cause the increase in chlorophyll content subsequentto the divisions of the chloroplast as demonstrated for Ulva pertusa during sporulation [26]. All clustersrepresent a marked changes of gene expression in the transition from determination phase (24 h) todifferentiation phase (48 h). 36 h after induction is regarded an important checkpoint duringgametogenesis after which blade cells are irreversibly committed to gametangium differentiation [15, 16,23]. Genes related to DNA replication and microtubule-based movement were signi�cantly up-regulatedafter this irreversible timepoint. Electron microscope observations indeed demonstrate drasticrearrangements in microtubule morphology between 36 h and 48 h after induction. The microtubulecytoskeleton of somatic cells consists of parallel microtubule bundles arranged mainly in the corticalcytoplasm parallel to the plasmalemma. However, 36 h after induction, the microtubule bundles traversethe cortical cytoplasm converging on a particular area and then the microtubules starting from a pointedsite at basal part of gametangium and forming a basket-like con�guration with a circular opening at thetop after 48 h [27]. Stratmann et al. (1996) analyzed the DNA synthesis during gametogenesis andconcluded that DNA replication starts from 25 hours after induction which is consistent with our RNA-seqdata. The enriched GO terms of microtubule-based movement in cluster 3 and 4 are the molecular basisof the cytological observation of the cytoskeleton organization. Genes in cluster 5 are upregulated onlyduring the swarming phase and the enriched GO terms mainly involve the synthesis of amino acid and

Page 9/22

isoprenoid pigment biosynthesis. Isoprenoid pigments biosynthetic pathway catalyzes the synthesis ofessential pigments of the photosystem structure in plant cells [28]. We speculate that the upregulation ofisoprenoid biosynthetic pathway is related to the reconstruction of the photosynthesis system for thegamete formation.

Transcription factors associated with gametogenesisTranscription factors play a pivotal role in the gene expression networks. TFs were surveyed in thedevelopmental program of gametogenesis that involves extensive cellular morphogenesis andsubsequent cell division and proliferation. Ulva gametogenesis was initiated by fragmentation andsubsequent removal of SI-1 and SI-2. As shown by Stratmann et al. (1996), fragmentation of growinggametophytes itself does not initiate gametogenesis and the cells remain in a "vegetative" state after 72 hbecause of the rapidly excreted SI-1 in the culture medium suppresses gametogenesis during thedetermination phase. To remove the effect of fragmentation and to identify a putative regulator initiatinggametogenesis in Ulva, we designed an experiment where thalli are fragmented but the sporulationinhibitors are not removed. With this control group, we identi�ed 12 up-regulated and 6 down-regulatedTFs at 6 h after induction of gametogenesis which constitute about 4–5% of the TF repertoire of U.mutabilis.

We selected the RWP-RK TF (UMSL057_0048) for further analysis because studies indicate a conservedrole of RWP-RK domain containing TFs in gametogenesis process [24, 29–31]. Nevertheless, the otherTFs identi�ed in this study might be potentially interesting as well. For example, Myb domain protein 98(MYB98, UMSL012_0085) is involved in the regulation of synergid differentiation in angiosperms [32, 33].Similarly, homologs of high nobility group transcription factor (UMSL024_0043) are relevant for thetransition of a vegetative to a reproductive phase [34–36]. The homolog of UMSL024_0043, which is themost up-regulated gene compared to the control group at 6 h, encodes SSRP1, which is component of thefacilitates chromatin transcription (FACT) complex in Arabidopsis. FACT is a conserved heterodimerichistone chaperone among eukaryotes and facilitates expression of FLC (FLOWERING LOCUS C) bybinding to the promoter of FLC, which adjusts the switch from vegetative to reproductive development inArabidopsis [37, 38]. Reduced amounts of SSRP1 result in decreased expression of FLC, thus accelerating�owering [38]. In contrast, UMSL024_0043 is up-regulated during gametogenesis and we did not identifya homolog of FLC in the Ulva genome. Thus, the speci�c function and the target genes of UMSL024_0043will need to be con�rmed experimentally.

One particular TF (encoded by UMSL057_0048) containing a RWP-RK protein motif that is important inDNA binding, is upregulated at 6h. Remarkably, minus dominance (MID), homolog of RWP-RK TFs in C.reinhardtii, has been described to be required for gametogenesis [39]. In addition, similar functions ofRWP-RK TFs during gametogenesis were reported in many other plants [24, 40]. RWP-RK transcriptionfactors are found throughout the Viridiplantae. Phylogenetic analysis divided the RWP-RK homologs into4 subfamilies, RKD(A), RKD(B), RKD(C) and NLP subfamily [29]. Previous studies on algae and higherplants showed that members of NLP subfamily were reported as the early regulators of cellular responseto N supply [41, 42] and RKD homologs plays evolutionarily conserved roles in germ cell differentiation [2,

Page 10/22

29, 31, 41–43]. Unlike other organisms which have multiple RWP-RK transcription factors, only two RWP-RK transcription factors are identi�ed in the U. mutabilis slender genome. A recent study on the Ulvapartita mating type locus structure reported three RWP-RK TFs in the U. partita genome, including oneRWP-RK TF located in the mating type minus locus with weak homology with MID gene [44]. The genomeapplied in our analysis is mating type plus, so the gene located in the mating type minus is notcharacterized. UMSL048_0014 contains a conserved GAF domain that is shared in the NLP subfamily.Two members of this subfamily were identi�ed as nitrogen response TF in C. reinhardtii (CreNIT2) andVolvox carteri (VcaNIT2) [41, 42]. Interestingly, the gametogenesis of C. reinhardtii is activated by the N-removal induction, while nitrogen concentration seems to have a positive effect on reproduction in Ulva[45]. This may suggest that UMSL048_0014 may have a function in nitrogen metabolism pathway only.UMSL057_0048 identi�ed in our study shares a conserved protein domain with RKD(A) family. In ourstudy, UMSL057_0048 is exclusively upregulated in the gametogenesis group (CW and NCW) and showsupregulation at 2 h after fragmentation and washing treatments. These �ndings indicate that RKD familymember UMSL057_0048 plays an essential role in Ulva gametogenesis.

Compared to UMSL057_0048, UMSL048_0014 contains a truncated GAF domain, which is conserved inthe NIN-like proteins (NLPs) subfamily and this subfamily is stated to be the early regulators of cellularresponse to N supply [29, 41, 42]. In contrast, UMSL057_0048 contains a conserved protein domain,which Chardin et al. referred to as motif 12, which is conserved in RKD subfamily (Fig. 6) and thissubfamily is involved in gametogenesis [2, 29, 41–43].

ConclusionsOur present work provides an important resource to study gametogenesis regulatory framework in amodel organism from which insights can be translated into commercially important Ulva species.Transcriptomic pro�ling helps us understand the gene regulatory networks during Ulva gametogenesis.The CNW group provides useful information to understand Ulva response to fragmentation andcombined fragmentation and washing factors respectively. The genetic transformation system isavailable in U. mutabilis and further genome editing technology is under development [46], which willallow functional characterisation of gametogenesis-related TFs in future experiments. The understandingof the mechanisms of inducing Ulva gametogenesis would be helpful to the commercial cultivation ofUlva, extending to other seaweeds. In addition, Ulva has developed multicellularity independently fromland plants and our study suggests that some TFs play a conserved role in reproduction throughout thegreen lineage. Therefore, further study of gametogenesis in Ulva should provide us with evolutionaryinsights into control of the sexual reproduction process for the Viridiplantae in general.

MethodsStrains and Culture Conditions

Page 11/22

Haploid gametophytes of Ulva mutabilis [strain 'wildtype' (wt-)] were collected initially in southernPortugal by B. Føyn [47]. During the last half century, U. mutabilis has been cultivated and permanentlypropagated under laboratory conditions and has been used as a model system for studying algaldevelopment by classical methods of plant physiology and genetics [6-7]. The strain used in this studywas acquired from the lab of Thomas Wichard, Friedrich Schiller University Jena in 2015 [15-16]. Andgametophytes of U. mutabilis were raised parthenogenetically from unmated gametes and culturedin transparent plastic boxes containing 2 L Provasoli enriched seawater (PES) medium [48] with thefollowing speci�c parameters: light intensity, 70 µmol photons m-2·s-1; temperature, 18 ± 1 °C; and 17:7h light:dark cycle in our lab. The medium was completely changed every 2 weeks until fertility (~10weeks). Afterwards, the medium was partially (20 %) changed to avoid induction of gametogenesis.

Induction of gametogenesis

Intact mature thalli of U. mutabilis were cut into 1-3 mm2 fragments using a herb chopper. Fragmentationwas carried out in the morning. Fragments derived from a single individual were separated into 2 groups.Fragments were washed 3 times for 15 minutes with autoclaved seawater and transferred to newmedium (CW, chopping and washing treatment). The washing step removes sporulation inhibitorsthereby inducing differentiation of the gametangia [15, 16]. Triplicates of 0.2 g fragmented thallus werecollected at 0, 6, 24, 48 and 72 h following gametogenesis induction. Control treatments consisted ofthalli that were fragmented but not washed and transferred immediately to the original culture medium(CNW, chopping no washing). The control was sampled at 6 and 24 h, respectively. The presence of SI-Iinhibits gametogenesis and control thalli therefore remain 'vegetative' after 72 h. A third and a fourthtreatment, used for expression analyses of RWP-RK transcription factors, consisted of thalli with normalgrowth that were neither chopped nor washed (NCNW) or not chopped but washed (NCW),respectively. Samples were �ash-frozen in liquid nitrogen and stored at -80 °C for RNA extraction. 

RNA extraction and RNAseq

Total RNA was extracted using a CTAB method [49]. Quality and quantity of total RNA was evaluatedusing a NanoDrop™ 2000c spectrophotometer (Thermo Fisher Scienti�c) and Bioanalyzer RNA6000(Agilent Technologies). cDNA libraries were constructed with a Quantseq™ 3' mRNA-Seq library prep kit(Lexogen) following the manufacturer's instructions. Illumina sequencing was performed using theIllumina Nextseq 500 platform to produce 50 bp single-end reads. RNA reads are available as SRA-xxx.Reads were mapped to the U. mutabilis genome [7]. We used the version of the slender strain which hasUTR-regions annotated, as opposed to the wildtype genome. Reads were mapped using TopHat ver.2.1.1 [50] with default parameters. The number of mapped reads was calculated using HTseq [51].Differentially expressed genes (DEG) were identi�ed between each treatment and control groups usingthe R package edgeR [52, 53]. The sample variation was estimated by tag-wise dispersion. Raw countswere normalized by CPM (counts per million reads mapped). Genes whose CPM value was >1 wereconsidered as 'expressed'. A false discovery rate (FDR) of 0.05 and absolute value of log 2-fold change of> 1 were adopted as thresholds for differential expressed genes (DEGs) detection. 

Page 12/22

Hierarchical Clustering Analysis

To study the transcriptional dynamics during the gametogenesis process, DEGs between any two of �vetime points (0 h, 6 h, 24 h, 48 h, 72 h) of the treatment group were identi�ed. The identi�ed DEGs weresubjected to a hierarchical clustering analysis by Pearson correlation [54] based on their CPM values inthe R (ver. 3.5.0) programming environment using the hclust function. Prior to the analysis, the optimalnumber of clusters was identi�ed and investigated by performing a SSE merit analysis and an R-basedaverage Silhouette Width analysis [55]. SSE is de�ned as the sum of the squared distance between eachcluster member and its centroid cluster. As the number of clusters increases, the distance between eachpoint and its centroid will be smaller. The optimal number of clusters is suggested when the addition of anew cluster does not signi�cantly decrease the SSE. The silhouette value describes how similar a gene isto its own cluster (cohesion) compared to other clusters (separation).

Functional Annotation and GO enrichment

Transcription factor (TF) identi�cation and GO terms for gene models were retrieved from the Ulvamutabilis genome annotation �le (downloaded from https://bioinformatics.psb.ugent.be/orcae/). The tophit of homologs acquired by the integrative orthology method of the differential expressed TFs of A.thaliana and C. reinhardtii were obtained using a custom-built PLAZA-Ulva version [56]. The functionalannotation of these homologs was acquired on TAIR (https://www.arabidopsis.org/) or Phytozome(https://phytozome.jgi.doe.gov/pz/portal.html). GO enrichment analysis for DEGs was based on aFisher's exact test implemented in the TopGO package in R [57]. The enriched GO terms (p<0.01) in thecategory "biological function" were summarized using the REVIGO web server, which performed aclustering algorithm that relies on the semantic similarity method [58].

TF identi�cation analysis

To verify RNA-seq results and to obtain more detailed expression pattern for speci�c TFs duringgametogenesis, we performed RT-qPCR analysis of the expression of up-regulated TFs. We sampled thetreatment group (CW, described above) and thalli with normal growth that were neither chopped norwashed (NCNW) every 2 h in the �rst 48 h during gametogenesis to obtain detailed expression pro�les forTFs identi�ed in the initial RNA-seq experiment. We designed an experiment consisting of four groups:normal vegetative thalli without chopping nor washing (NCNW); chopped thalli that were not washed(CNW); chopped and washed thalli (CW, normal induction); intact thalli that were washed to remove of SIin the medium (NCW). Gametogenesis was partly induced in the latter treatment. Thalli were sampled at 6h for the four treatments to identify the expression pro�le of the TFs. RNA was extracted as describedabove and cDNA was synthesized using an iScriptTM cDNA Synthesis Kit (Bio-Rad). PCR was performedusing Bio-Rad CFX96 Real-Time PCR systems. Reactions were performed in a �nal volume of 10 μLcontaining 5 μL of SYBR Green Master Mix, 0.5 μM each primer, and 10 ng of cDNA under followingprogram: 5 min at 95 °C followed by 35 cycles of 95 °C for 10 s and 55 °C for 30 s. Gene expression levelswere calculated using the ddCt method [59]. UBIQUITIN (UMSL012_0173) and ELFA (UMSL 016_0119)

Page 13/22

were stable under a wide range of time points and were taken as reference. Primer sequences are listed inTable S1.

DeclarationsEthics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Data availability

The raw reads generated in this study are available on SRAxxx. The data can not be made public for thetime being but are available from the corresponding author on reasonable request.

Competing interests

The authors declare that they have no competing interests. 

Funding

This work makes use of resources and facilities provided by UGent as part of the Belgian contribution toEMBRC-ERIC (FWO GOH3817N) and the Betty Moore Foundation for a SYMBIOSIS grant (nr.4546891618). X.L. is indebted to the China Scholarship Council (201504910698) and Ghent UniversitySpecial Research Fund (BOF-16/CHN/023) for a PhD grant and to COST Action FA1406 (Phycomorph)for a Short-Term Scienti�c Mission to Friedrich Schiller University Jena. J.B. thanks the ResearchFoundation – Flanders (FWO) for a postdoctoral fellowship (12T3418N) and Ghent University SpecialResearch Grant (BOF20/PDO/016). T.W. was funded by the Deutsche Forschungsgemeinschaft throughGrant No. SFB 1127/2 ChemBioSys–239748522. 

Author contributions

X.L., and O.D.C. designed the study. X.L., T.W. and S.D. performed experiments. X.L., J.B. and K.B.performed the transcriptional analysis. X.L. wrote the manuscript with contributions from all authors.

Acknowledgements

Not applicable.

References

Page 14/22

1. Coelho SM, Peters AF, Charrier B, Roze D, Destombe C, Valero M, Cock JM: Complex life cycles ofmulticellular eukaryotes: new approaches based on the use of model organisms. Gene 2007, 406(1–2):152–170.

2. Lin H, Goodenough UW: Gametogenesis in the Chlamydomonas reinhardtii minus mating type iscontrolled by two genes, MID and MTD1. Genetics 2007, 176(2):913–925.

3. Goodenough U, Lin H, Lee JH: Sex determination in Chlamydomonas. Semin Cell Dev Biol 2007,18(3):350–361.

4. Abe J, Kubo T, Takagi Y, Saito T, Miura K, Fukuzawa H, Matsuda Y: The transcriptional program ofsynchronous gametogenesis in Chlamydomonas reinhardtii. Curr Genet 2004, 46(5):304–315.

5. Abe J, Kubo T, Saito T, Matsuda Y: The regulatory networks of gene expression during the sexualdifferentiation of Chlamydomonas reinhardtii, as analyzed by mutants for gametogenesis. Plant CellPhysiol 2005, 46(2):312–316.

�. Wichard T, Charrier B, Mineur F, Bothwell JH, Clerck OD, Coates JC: The green seaweed Ulva: a modelsystem to study morphogenesis. Front Plant Sci 2015, 6:72.

7. De Clerck O, Kao S-M, Bogaert KA, Blomme J, Fo�onker F, Kwantes M, Vancaester E, VanderstraetenL, Aydogdu E, Boesger J: Insights into the evolution of multicellularity from the sea lettuce genome.Curr Biol 2018, 28(18):2921–2933. e2925.

�. Balar NB, Mantri VA: Insights into life cycle patterns, spore formation, induction of reproduction,biochemical and molecular aspects of sporulation in green algal genus Ulva: implications forcommercial cultivation. J Appl Phycol 2020, 32(1):473–484.

9. Nisizawa K, Noda H, Kikuchi R, Watanabe T: The main seaweed foods in Japan. Hydrobiologia 1987,151(1):5–29.

10. Neori A, Chopin T, Troell M, Buschmann AH, Kraemer GP, Halling C, Shpigel M, Yarish C: Integratedaquaculture: rationale, evolution and state of the art emphasizing seaweed bio�ltration in modernmariculture. Aquaculture 2004, 231(1–4):361–391.

11. Alsufyani T, Engelen AH, Diekmann OE, Kuegler S, Wichard T: Prevalence and mechanism ofpolyunsaturated aldehydes production in the green tide forming macroalgal genus Ulva (Ulvales,Chlorophyta). Chem Phys Lipids 2014, 183:100–109.

12. Bolton JJ, Cyrus MD, Brand MJ, Joubert M, Macey BM: Why grow Ulva? Its potential role in the futureof aquaculture. Perspectives in Phycology 2016, 3(3):113–120.

13. Wang Y, Liu F, Liu X, Shi S, Bi Y, Moejes FW: Comparative transcriptome analysis of four co-occurringUlva species for understanding the dominance of Ulva prolifera in the Yellow Sea green tides. J ApplPhycol 2019, 31(5):3303–3316.

14. Lüning K, Kadel P, Pang S: Control of reproduction rhythmicity by environmental and endogenoussignals in Ulva Pseudocurvata (Chlorophyta). J Phycol 2008, 44(4):866–873.

15. Stratmann J, Paputsoglu G, Oertel W: Differentiation of Ulva Mutabilis (Chlorophyta) gametangiaand gamete release are controlled by extracellular inhibitors. J Phycol 1996, 32:1009–1021.

Page 15/22

1�. Wichard T, Oertel W: Gametogenesis and gamete release of Ulva mutabilis and Ulva lactuca(Chlorophyta): regulatory effects and chemical characterization of the “Swarming Inhibitor”. J Phycol2010, 46(2):248–259.

17. Sousa AI, Martins I, Lillebø AI, Flindt MR, Pardal MA: In�uence of salinity, nutrients and light on thegermination and growth of Enteromorpha sp. spores. Journal of Experimental Marine Biology andEcology 2007, 341(1):142–150.

1�. Mantri VA, Singh RP, Bijo A, Kumari P, Reddy C, Jha B: Differential response of varying salinity andtemperature on zoospore induction, regeneration and daily growth rate in Ulva fasciata (Chlorophyta,Ulvales). J Appl Phycol 2011, 23(2):243–250.

19. Hiraoka M, Enomoto S: The induction of reproductive cell formation of Ulva pertusa Kjellman(Ulvales, Ulvophyceae). Phycol Res 1998, 46(3):199–203.

20. Zhang X, Xu D, Mao Y, Li Y, Xue S, Zou J, Lian W, Liang C, Zhuang Z, Wang Q: Settlement ofvegetative fragments of Ulva prolifera con�rmed as an important seed source for succession of alarge-scale green tide bloom. Limnol Oceanogr 2011, 56(1):233–242.

21. Vesty EF, Kessler RW, Wichard T, Coates JC: Regulation of gametogenesis and zoosporogenesis inUlva linza (Chlorophyta): comparison with Ulva mutabilis and potential for laboratory culture. FrontPlant Sci 2015, 6:15.

22. Gao S, Chen X, Yi Q, Wang G, Pan G, Lin A, Peng G: A strategy for the proliferation of Ulva prolifera,main causative species of green tides, with formation of sporangia by fragmentation. PLoS ONE2010, 5(1).

23. Kessler RW, Crecelius AC, Schubert US, Wichard T: In situ monitoring of molecular changes duringcell differentiation processes in marine macroalgae through mass spectrometric imaging. Analyticaland Bioanalytical Chemistry 2017, 409(20):4893–4903.

24. Tedeschi F, Rizzo P, Rutten T, Altschmied L, Baumlein H: RWP-RK domain-containing transcriptionfactors control cell differentiation during female gametophyte development in Arabidopsis. NewPhytol 2017, 213(4):1909–1924.

25. He Y, Wang Y, Hu C, Sun X, Li Y, Xu N: Dynamic metabolic pro�les of the marine macroalga Ulvaprolifera during fragmentation-induced proliferation. PLoS ONE 2019, 14(5).

2�. Park J: Photosynthetic and biochemical traits change in the green-tide‐forming macroalga Ulvapertusa during sporulation. J Phycol 2020, 56(2):549–557.

27. Katsaros C, Weiss A, Llangos I, Theodorou I, Wichard T: Cell structure and microtubule organisationduring gametogenesis of Ulva mutabilis Føyn (Chlorophyta). Bot Mar 2017, 60(2).

2�. Couso I, Vila M, Vigara J, Cordero BF, Vargas MÁ, Rodríguez H, León R: Synthesis of carotenoids andregulation of the carotenoid biosynthesis pathway in response to high light stress in the unicellularmicroalga Chlamydomonas reinhardtii. Eur J Phycol 2012, 47(3):223–232.

29. Chardin C, Girin T, Roudier F, Meyer C, Krapp A: The plant RWP-RK transcription factors: keyregulators of nitrogen responses and of gametophyte development. J Exp Bot 2014, 65(19):5577–5587.

Page 16/22

30. Rovekamp M, Bowman JL, Grossniklaus U: Marchantia MPRKD regulates the gametophyte-sporophyte transition by keeping egg cells quiescent in the absence of fertilization. Curr Biol 2016,26(13):1782–1789.

31. Yin M, Zhang Z, Xuan M, Feng H, Ye W, Zheng X, Wang Y: Conserved subgroups of the plant-speci�cRWP-RK transcription factor family are present in Oomycete Pathogens. Front Microbiol 2020,11:1724.

32. Kasahara RD, Portereiko MF, Sandaklie-Nikolova L, Rabiger DS, Drews GN: MYB98 is required forpollen tube guidance and synergid cell differentiation in Arabidopsis. The Plant Cell 2005,17(11):2981–2992.

33. Punwani JA, Rabiger DS, Drews GN: MYB98 positively regulates a battery of synergid-expressedgenes encoding �liform apparatus–localized proteins. The Plant Cell 2007, 19(8):2557–2568.

34. Klepikova AV, Logacheva MD, Dmitriev SE, Penin AA: RNA-seq analysis of an apical meristem timeseries reveals a critical point in Arabidopsis thaliana �ower initiation. BMC Genomics 2015,16(1):466.

35. Searle I, He Y, Turck F, Vincent C, Fornara F, Kröber S, Amasino RA, Coupland G: The transcriptionfactor FLC confers a �owering response to vernalization by repressing meristem competence andsystemic signaling in Arabidopsis. Genes Dev 2006, 20(7):898–912.

3�. Pfab A, Grønlund JT, Holzinger P, Längst G, Grasser KD: The Arabidopsis histone chaperone FACT:role of the HMG-box domain of SSRP1. J Mol Biol 2018, 430(17):2747–2759.

37. Grasser KD: The FACT histone chaperone: tuning gene transcription in the chromatin context tomodulate plant growth and development. Front Plant Sci 2020, 11(85).

3�. Lolas IB, Himanen K, Grønlund JT, Lynggaard C, Houben A, Melzer M, Van Lijsebettens M, GrasserKD: The transcript elongation factor FACT affects Arabidopsis vegetative and reproductivedevelopment and genetically interacts with HUB1/2. The Plant Journal 2010, 61(4):686–697.

39. Ferris PJ, Pavlovic C, Fabry S, Goodenough UW: Rapid evolution of sex-related genes in<bi>Chlamydomonas</bi>. Proceedings of the National Academy of Sciences 1997, 94(16):8634–8639.

40. Waki T, Hiki T, Watanabe R, Hashimoto T, Nakajima K: The Arabidopsis RWP-RK protein RKD4triggers gene expression and pattern formation in early embryogenesis. Curr Biol 2011, 21(15):1277–1281.

41. Camargo A, Llamas A, Schnell RA, Higuera JJ, Gonzalez-Ballester D, Lefebvre PA, Fernandez E,Galvan A: Nitrate signaling by the regulatory gene NIT2 in Chlamydomonas. Plant Cell 2007,19(11):3491–3503.

42. Castaings L, Camargo A, Pocholle D, Gaudon V, Texier Y, Boutet-Mercey S, Taconnat L, Renou JP,Daniel‐Vedele F, Fernandez E: The nodule inception‐like protein 7 modulates nitrate sensing andmetabolism in Arabidopsis. The Plant Journal 2009, 57(3):426–435.

43. Koi S, Hisanaga T, Sato K, Shimamura M, Yamato KT, Ishizaki K, Kohchi T, Nakajima K: Anevolutionarily conserved plant RKD factor controls germ cell differentiation. Curr Biol 2016,

Page 17/22

26(13):1775–1781.

44. Yamazaki T, Ichihara K, Suzuki R, Oshima K, Miyamura S, Kuwano K, Toyoda A, Suzuki Y, Sugano S,Hattori M: Genomic structure and evolution of the mating type locus in the green seaweed Ulvapartita. Sci Rep 2017, 7(1):11679.

45. Gao G, Clare AS, Rose C, Caldwell GS: Eutrophication and warming-driven green tides (Ulva rigida)are predicted to increase under future climate change scenarios. Mar Pollut Bull 2017, 114(1):439–447.

4�. Oertel W, Wichard T, Weissgerber A: Transformation of Ulva mutabilis (Chlorophyta) by vectorplasmids integrating into the genome. J Phycol 2015, 51(5):963–979.

47. Føyn B: Über die Sexualität und den Generationswechsel von Ulva mutabilis. Arch Protistenkd 1958,102:473–480.

4�. Provasoli L: Media and prospects for the cultivation of marine algae. In: Cultures and Collections ofAlgae Proceedings of US-Japan Conference, Hakone, September 1966: 1968. Japan Society of PlantPhysiology.

49. Le Bail A, Dittami SM, de Franco P-O, Rousvoal S, Cock MJ, Tonon T, Charrier B: Normalisation genesfor expression analyses in the brown alga model Ectocarpus siliculosus. BMC Mol Biol 2008, 9(1):75.

50. Trapnell C, Salzberg SL: How to map billions of short reads onto genomes. Nat Biotechnol 2009,27(5):455.

51. Anders S, Pyl PT, Huber W: HTSeq—a Python framework to work with high-throughput sequencingdata. Bioinformatics 2015, 31(2):166–169.

52. Robinson MD, McCarthy DJ, Smyth GK: edgeR: a Bioconductor package for differential expressionanalysis of digital gene expression data. Bioinformatics 2010, 26(1):139–140.

53. Team RC: R: A language and environment for statistical computing. 2013.

54. Corpet F: Multiple sequence alignment with hierarchical clustering. Nucleic Acids Res 1988,16(22):10881–10890.

55. Langfelder P, Zhang B, Horvath S: De�ning clusters from a hierarchical cluster tree: the Dynamic TreeCut package for R. Bioinformatics 2008, 24(5):719–720.

5�. Vandepoele K, Van Bel M, Richard G, Van Landeghem S, Verhelst B, Moreau H, Van de Peer Y,Grimsley N, Piganeau G: pico-PLAZA, a genome database of microbial photosynthetic eukaryotes.Environ Microbiol 2013, 15(8):2147–2153.

57. Alexa A, Rahnenführer J: Gene set enrichment analysis with topGO.https://bioconductororg/packages/release/bioc/vignettes/topGO/inst/doc/topGOpdf 2009.

5�. Supek F, Bošnjak M, Škunca N, Šmuc T: REVIGO Summarizes and Visualizes Long Lists of GeneOntology Terms. PLoS ONE 2011, 6(7):e21800.

59. Livak KJ, Schmittgen TD: Analysis of relative gene expression data using real-time quantitative PCRand the 2–∆∆CT method. Methods 2001, 25(4):402–408.

Page 18/22

Figures

Figure 1

Time course of gametogenesis in Ulva mutabilis. "determination phase" and "differentiation phase"described previously by Stratmann et al. (1996). The "swarming phase" is the time when the change ofmedium can release the gametes. Asterisks indicate the sampled time points for RNA-seq. (a) vegetativecells before induction, (b) gametogenesis induction after 24 h with chloroplast reorientation, (c)gametangia after 48 h, (d) reproductive cells with fully formed gametes and discharged gametes after 72h, (e) Ulva thallus before induction, (f) released gametes, (g,h) induced Ulva tissue without refreshment ofgrowth medium after 48 (g) and 72 hours (h), demonstrating that no gametes were formed as long assporulation inhibitors are not removed. Scale bars: 2cm (e) or 20 μm (others).

Page 19/22

Figure 2

Clusters of differentially expressed genes (DEG) identi�ed during the different phases of gametogenesisat 0, 6, 24, 48 and 72h. Clusters are presented according their maximal expression in the different phasesof gametogenesis. A selection of the most relevant overrepresented GO terms (p<0.01 and summarizedby REVIGO) is presented for each cluster. We refer to Table S5 for the full list enriched GO terms.

Page 20/22

Figure 3

Differentially expressed genes (DEGs) between fully induced thalli (CW) and thalli with the sporulationinhibitors still present (CNW). The Venn diagram depicts the number of up- and downregulated DEGs at 6h and 24 h.

Page 21/22

Figure 4

Enriched GO terms associated with DEGs between fully induced thalli (CW) and thalli with the sporulationinhibitors still present (CNW) at 6 h and 24 h respectively.

Page 22/22

Figure 5

Expression of RWP-RK transcription factor UMSL057_0048 during gametogenesis. (a). RT-qPCR analysisof UMSL057_0048 expression level compared to the control group (NCNW) in the �rst 48 h ofgametogenesis. Three biological replicates were used for analysis and the gene expression level wascalculated using the standard ddCt method. (b). Expression level of UMSL057_0048 at 6 h in induced(CW and NCW) versus not induced (CNW and NCNW) thalli. Three biological replicates were used for theexperiment and standard ddCt method was applied for analysis. Signi�cant differences are indicatedwith lowercase (ANOVA test, P < 0.05). Stand errors are indicated with errors bars.

Figure 6

Comparison of domain organization between RWP-RK proteins. “Motif 12” refers to the same motifidenti�ed by Chardin et al. (2014). Model organisms in which the RWP-RK protein were functionallyvalidated were chosen as reference. UMSL, Ulva mutabilis slender (+); Mp, Marchantia polymorpha; At,Arabidopsis thaliana; Cre, Chlamydomonas reinhardtii; Vca, Volvox carteri; Amino acid sequences andaccession numbers are listed in Table S6.

Supplementary Files

This is a list of supplementary �les associated with this preprint. Click to download.

Supplementary�guresS1S2.docx

SupplementarytablesS1S6.xlsx